Abstract
The relations existing at a stationary boundary between saturated and antisaturated magnetic phases in a stretched NiFe wire have been investigated and a theory has been developed. Experimentally, both the minimum reversed (stopping) field required to stop the propagation of a moving discontinuity as well as the flux and magnetic pole distribution in the stopped boundary have been determined. Over the major portion of the boundary the pole field subtracted from the main field, and this demagnetizing field was found just to neutralize the excess of the applied field over the critical field. This confirmed the view that reversal in the wire proceeds only while the totoal field exceeds the critical field. A simple pole strength distribution which complies with this view was found to be one in which pole strength varies linearly with distance. Its demagnetizing field is approximately constant and checks the known excess fields reasonably well. Further refinement followed from recognizing that at the forward end of the boundary the pole field adds to the main field and is neutralized there by the stopping field. As a result, it was possible, given the excess field, the length of stopping coil, and the total change in induction, to calculate roughly the length of the boundary and the magnitude of the minimum stopping field. Certain details of the pole and field distribution under the stopping coil have been considered, particularly the expected sharp increase in minimum stopping field as the main field approached the starting field for the wire. When first the main field and then the stopping field was removed, the boundary was found to remain in place without any major change. The theory accounts for the escape of such a stopped boundary when the magnetic fields were reapplied. Symmetrical regions of reversed magnetization have been formed, both "spindles" and "rings," and some features of their behavior have been explained.